Update README.md
Browse files
README.md
CHANGED
|
@@ -86,22 +86,9 @@ Not all prompts are equal. A factual recall question ("What is the capital of Fr
|
|
| 86 |
|
| 87 |
**Brick** solves this by routing each query to the right model tier in real time. Complexity classification is one of several routing signals (alongside keyword matching, domain detection, and reasoning-depth estimation) that Brick uses to make sub-50ms routing decisions.
|
| 88 |
|
| 89 |
-
|
| 90 |
-
|
| 91 |
-
|
| 92 |
-
β β
|
| 93 |
-
β ββββββββββββββββββ β βββββββββββββββββββ
|
| 94 |
-
β β Complexity ββββΌβββββΆβ easy β Qwen 7B β
|
| 95 |
-
β β Extractor β β β mediumβ Llama 70Bβ
|
| 96 |
-
β β (this model) β β β hard β Claude β
|
| 97 |
-
β ββββββββββββββββββ β βββββββββββββββββββ
|
| 98 |
-
β ββββββββββββββββββ β
|
| 99 |
-
β β Domain Det. β β
|
| 100 |
-
β β Keyword Match β β
|
| 101 |
-
β β Reasoning Est. β β
|
| 102 |
-
β ββββββββββββββββββ β
|
| 103 |
-
ββββββββββββββββββββββββ
|
| 104 |
-
```
|
| 105 |
|
| 106 |
## Model Details
|
| 107 |
|
|
@@ -257,8 +244,8 @@ model_pools:
|
|
| 257 |
|
| 258 |
### β οΈ Out-of-Scope Uses
|
| 259 |
- **Content moderation or safety filtering** β this model classifies cognitive difficulty, not content safety
|
| 260 |
-
- **Non-English queries**
|
| 261 |
-
- **Direct use as a chatbot or generative model**
|
| 262 |
|
| 263 |
## Limitations
|
| 264 |
|
|
@@ -315,7 +302,7 @@ Regolo.ai is committed to sustainable AI. This model was trained on GPU infrastr
|
|
| 315 |
|
| 316 |
## About Regolo.ai
|
| 317 |
|
| 318 |
-
[Regolo.ai](https://regolo.ai) is the EU-sovereign LLM inference platform built on [Seeweb](https://www.seeweb.it/) infrastructure. We provide zero-data-retention, GDPR-native AI inference for enterprises that need privacy, compliance, and performance
|
| 319 |
|
| 320 |
**Brick** is our open-source semantic routing system that intelligently distributes queries across model pools, optimizing for cost, latency, and quality.
|
| 321 |
|
|
|
|
| 86 |
|
| 87 |
**Brick** solves this by routing each query to the right model tier in real time. Complexity classification is one of several routing signals (alongside keyword matching, domain detection, and reasoning-depth estimation) that Brick uses to make sub-50ms routing decisions.
|
| 88 |
|
| 89 |
+
|
| 90 |
+
<img src="https://cdn-uploads.huggingface.co/production/uploads/66e9a629df006ca4588b82bd/ZoRBcn8rD8sTEHdkiczOO.png" alt="brick_router" width="800">
|
| 91 |
+
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 92 |
|
| 93 |
## Model Details
|
| 94 |
|
|
|
|
| 244 |
|
| 245 |
### β οΈ Out-of-Scope Uses
|
| 246 |
- **Content moderation or safety filtering** β this model classifies cognitive difficulty, not content safety
|
| 247 |
+
- **Non-English queries** trained on English data only; accuracy degrades significantly on other languages
|
| 248 |
+
- **Direct use as a chatbot or generative model** this is a classification adapter, not a generative model
|
| 249 |
|
| 250 |
## Limitations
|
| 251 |
|
|
|
|
| 302 |
|
| 303 |
## About Regolo.ai
|
| 304 |
|
| 305 |
+
[Regolo.ai](https://regolo.ai) is the EU-sovereign LLM inference platform built on [Seeweb](https://www.seeweb.it/) infrastructure. We provide zero-data-retention, GDPR-native AI inference for enterprises that need privacy, compliance, and performance all from European data centers powered by renewable energy.
|
| 306 |
|
| 307 |
**Brick** is our open-source semantic routing system that intelligently distributes queries across model pools, optimizing for cost, latency, and quality.
|
| 308 |
|